A description of Google’s strategy and internal culture from its “adult supervisor,” Eric Schmidt. I was surprised to read some of the highlights below, including a pretty spirited argument against the practice of telecommuting. This is especially worth reading if you have experience working in big tech companies – there are probably a lot of things your company does differently from Google and it’s interesting to pick them out and ask why.


In a normal company, the CEO, seeing a bad product, would call the person in charge of the product. There would be a meeting or two or three to discuss the problem, review potential solutions, and decide on a course of action. A plan would come together to implement the solution. Then, after a fair amount of quality assurance testing, the solution would launch. In a normal company, this would take several weeks. This isn’t what Larry did. Instead, he printed out the pages containing the results he didn’t like, highlighted the offending ads, posted them on a bulletin board on the wall of the kitchen by the pool table, and wrote THESE ADS SUCK in big letters across the top. Then he went home.
When you can reach out and tap someone on the shoulder, there is nothing to get in the way of communications and the flow of ideas. The traditional office layout, with individual cubicles and offices, is designed so that the steady state is quiet. Most interactions between groups of people are either planned (a meeting in a conference room) or serendipitous (the hallway / water cooler / walking through the parking lot meeting). This is exactly backward; the steady state should be highly interactive, with boisterous, crowded offices brimming with hectic energy.
We invest in our offices because we expect people to work there, not from home. Working from home during normal working hours, which to many represents the height of enlightened culture, is a problem that—as Jonathan frequently says—can spread throughout a company and suck the life out of its workplace.
When it comes to the quality of decision-making, pay level is intrinsically irrelevant and experience is valuable only if it is used to frame a winning argument. Unfortunately, in most companies experience is the winning argument. We call these places “tenurocracies,” because power derives from tenure, not merit. It reminds us of our favorite quote from Jim Barksdale, erstwhile CEO of Netscape: “If we have data, let’s look at data. If all we have are opinions, let’s go with mine.”
For a meritocracy to work, it needs to engender a culture where there is an “obligation to dissent.”
We’ve worked at other companies with a rule of seven, but in all of those cases the rule meant that managers were allowed a maximum of seven direct reports. The Google version suggests that managers have a minimum of seven direct reports (Jonathan usually had fifteen to twenty when he ran the Google product team). We still have formal organization charts, but the rule (which is really more of a guideline, since there are exceptions) forces flatter charts with less managerial oversight and more employee freedom. With that many direct reports—most managers have a lot more than seven—there simply isn’t time to micromanage.
Once you identify the people who have the biggest impact, give them more to do. When you pile more responsibility on your best people, trust that they will keep taking it on or tell you when enough is enough. As the old saying goes: If you want something done, give it to a busy person.
(Eric was once asked at a company meeting what the Google dress code was. “You must wear something” was his answer.)
(TV star Tina Fey has her own version of the LAX test, which she credits to Saturday Night Live producer Lorne Michaels: “Don’t hire anyone that you wouldn’t want to run into by the bathrooms at three in the morning, because you’re going to be [in the office] all night.”)
This is why most conference rooms at Google have two projectors. One of them is for videoconferencing with other offices or for projecting meeting notes. The other is for data. When discussing options and opinions, we start the meetings with data. We don’t seek to convince by saying “I think.” We convince by saying “Let me show you.”
The company’s intranet, Moma, includes information on just about every upcoming product, for example, and the weekly TGIF meeting usually features presentations by product teams of coming attractions, complete with demos and screenshots of cool things that are under development. Attending TGIF would be like scoring a Willy Wonka golden ticket for those legions of bloggers who love to speculate on what Google will be up to next, because we put it all out there, sharing stuff that in most companies would stay carefully hidden. Again, no grainy photos of screenshots or herky-jerky videos of demos are leaked, furtively captured from the back row. We trust our employees with all sorts of vital information, and they honor that trust.
OKRs are another great example of transparency. These are an individual’s Objectives (the strategic goals to accomplish) and Key Results (the way in which progress toward that goal is measured). Every employee updates and posts his OKRs company-wide every quarter, making it easy for anyone to quickly find anyone else’s priorities. When you meet someone at Google and want to learn more about what they do, you go on Moma and read their OKRs. This isn’t just a job title and description of the role, it’s their first-person account of the stuff they are working on and care about. It’s the fastest way to figure out what motivates them.
Snippets are like weekly status reports that cover a person’s most important activities for a week, but in a short, pithy format, so they can be written in just a few minutes or compiled (in a doc or draft email) as the week goes on. There is no set format, but a good set of snippets includes the most important activities and achievements of the week and quickly conveys what the person is working on right now, from cryptic (“SMB Framework,” “10% list”) to mundane (“completed quarterly performance reviews,” “started family vacation”). Like OKRs, they are shared with everyone. Snippets are posted on Moma, where anyone can see anyone else’s, and for years Larry received a weekly compendium of the snippets from engineering and product leads. That way he always could get at the truth.
Knowledge Graph, which launched in 2012, populates the right-hand side of the web results page for queries about people, places, and things with a concisely formatted panel of algorithmically curated information about that entity. It pulls all the most relevant facts together into one, easy-to-read box. For most queries, the panel replaces the ads that previously appeared on that part of the page.
At Google, our users are the people who use our products, while our customers are the companies that buy our advertising and license our technology. There are rarely conflicts between the two, but when there are, our bias is toward the user. It has to be this way, regardless of your industry.
While the 70/20/10 rule ensured that our core business would always get the bulk of the resources and that the promising, up-and-coming areas would also get investment, it also ensured that the crazy ideas got some support too, and were protected from the inevitable budget cuts. And 10 percent isn’t a lot of resources, which is fine, because overinvesting in a new concept is just as problematic as underinvesting, since it can make it much harder to admit failure later on. Million-dollar ideas are a lot harder to kill than thousand-dollar ones, so overinvestment can create a situation wherein willful confirmation bias—the tendency to see only the good things in projects in which a lot has been invested—obscures sound decision-making.
In 2002, Larry Page began wondering if it was possible to make every book ever published searchable online—not just the most popular titles or some other subset, but every single book. (We later calculated there were precisely 129,864,880 different book titles in the world.)177 When every book ever published was available online, he reasoned, and when universal translation became available, then all of the world’s knowledge would be accessible to every person. As the cofounder, Larry could have assigned a team of engineers to the problem and given them a nice budget. Instead he got a digital camera, rigged it to a tripod, and set the contraption up on a table in his office. He pointed the camera down at the table, turned on a metronome to pace his movements, and started snapping pictures while Marissa Mayer turned the pages. Based on this crude prototype, they were able to estimate what it took to digitize a book, and made some calculations that the audacious project was indeed feasible. Google Books was born.
When you want to spur innovation, the worst thing you can do is overfund it.
In fact, 20 percent time is more like 120 percent time, since it often occurs on nights and weekends. But it can also be stored up and used all at once—Jonathan had one product manager take a summer to work on a 20 percent project. Regardless of when you take your 20 percent time, assuming it doesn’t get in the way of doing your regular job, no one can stop you from doing it. Twenty percent time is a check and balance on imperial managers, a way to give people permission to work on stuff they aren’t supposed to work on. It helps bring to life the Steve Jobs maxim that “you have to be run by ideas, not hierarchy.”
The most valuable result of 20 percent time isn’t the products and features that get created, it’s the things that people learn when they try something new. Most 20 percent projects require people to practice or develop skills outside of those they use on a day-to-day basis, often collaborating with colleagues they don’t regularly work with.